perf: Do not get values from disabled counters in group format read
authorJiri Olsa <jolsa@redhat.com>
Mon, 15 Oct 2012 18:13:45 +0000 (20:13 +0200)
committerArnaldo Carvalho de Melo <acme@redhat.com>
Wed, 7 Aug 2013 20:35:19 +0000 (17:35 -0300)
It's possible some of the counters in the group could be
disabled when sampling member of the event group is reading
the rest via PERF_SAMPLE_READ sample type processing. Disabled
counters could then produce wrong numbers.

Fixing that by reading only enabled counters for PERF_SAMPLE_READ
sample type processing.

Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/n/tip-wwkjb0bbcuslnz0klrmqi26r@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
kernel/events/core.c

index 5200b608b4818a7a80887a5127dc276b96200ce0..e82e70025d4233ea67cbb9b3ae280f759896ba09 100644 (file)
@@ -4388,7 +4388,8 @@ static void perf_output_read_group(struct perf_output_handle *handle,
        list_for_each_entry(sub, &leader->sibling_list, group_entry) {
                n = 0;
 
-               if (sub != event)
+               if ((sub != event) &&
+                   (sub->state == PERF_EVENT_STATE_ACTIVE))
                        sub->pmu->read(sub);
 
                values[n++] = perf_event_count(sub);